MIMO Array Imaging for Ultrasonic Nondestructive Testing

Ultrasonic sensor arrays continue to be broadly applied for nondestructive material testing. Generally, conventional beamforming techniques have been the favorite approach to generate images from the sensor array data. In this paper, we examine the use of multiple-input multiple-output (MIMO) ultrasonic processing technique for imaging internal structures of materials. The goal is to identify and locate potential defects and anomalies. The imaging technique is comprised of excitation of transmitting sensors with sequential or orthogonal wideband signals, matched filtering, and adaptive weighting. The weighting of the signals at the receiver takes into account the transducer ultrasound radiation patterns. The MIMO technique is particularly attractive for ultrasonic imaging, as the different bistatic combinations of transmit and receive sensor pairs allows effective and simple formations of virtual arrays with extended apertures and denser spatial sampling. As such, high-resolution images can be generated with fewer or available transducers. The performance of this technique is experimentally examined using test specimens with artificially drilled small size flat bottom holes that simulate defects. One-dimensional and two-dimensional array configurations are used to form desired virtual arrays and their respective imaging capabilities are evaluated and compared.